|Home | About | Journals | Submit | Contact Us | Français|
The capacity for robust proliferation upon re-infection is a hallmark of adaptive immunity and the basis of vaccination. A widely used animal model for the study of human disease is the rhesus macaque (RM), where capacity for proliferation can be assessed ex vivo using carboxyfluorescein succinimidyl ester (CFSE)-based dilution assays. However, we show over the course of the standard ex vivo proliferation assay that CFSE-labeling at commonly-used dye concentrations induces significant cell death, but that this phenomenon is dose-dependent. Here we describe an alternative, semi-quantitative method for estimating T cell proliferative responses that avoids the putative biases associated with chemical modification. RM peripheral blood mononuclear cells were stimulated ex vivo with cognate peptides for five days, immunostained for intracellular Ki-67, and then analyzed by flow cytometry. We describe a gating strategy using Ki-67 and side light scatter, also a marker of blastogenesis, which correlates strongly with data from CFSE dilution. We show that this method is a valid tool for measuring RM antigen-specific cellular proliferation ex vivo and can be used as an alternative to CFSE dilution assays.
The ability of memory T cells to mount robust effector responses upon antigenic restimulation is a hallmark of immunity and the basis of vaccination (1,2). Assessment of cell proliferation after antigenic or mitogenic stimulation is an important factor in clinical and research laboratories for the diagnosis of immunodeficiencies and the evaluation of vaccine efficacy (3), respectively. A widely used animal model for the study of human disease is the rhesus macaque (RM) (4), where capacity for proliferation can be assessed by ex vivo assays designed to measure T cell proliferation; the incorporation of tritiated thymidine (3H) or a thymidine analog, bromodeoxyuridine (BrdU), into lymphocyte DNA can be used to measure proliferative responses, as well as the dilution of cellular dyes such as carboxyfluorescein succinimidyl ester (CFSE) (5), the level of which decreases upon cellular division (6,7). To date, the uptake of radioactive 3H has been the gold standard due to its accuracy and sensitivity, but has been challenged by the popular, nonradioactive CFSE-based dilution assay. This assay is now a widely accepted, nonradioactive alternative and works when the carboxyfluorescein diacetate, succinimidyl ester (CFDA-SE)-dye passively diffuses into cells where its acetate groups are cleaved by intracellular esterases to yield highly fluorescent, amine-reactive CFSE (CFSE is not cell permeable). The succinimidyl ester group reacts with intracellular amines, forming fluorescent dye-protein adducts that are well-retained by the cells throughout development, mitosis, and experimental tracing. This method allows for examination of specific populations of proliferating cells and identifies up to 7 to 10 successive cell generations and dye can sometimes last up to eight weeks in mice upon adoptive transfer (8). The sequential halving of fluorescence is visualized as distinct peaks when analyzed by flow cytometry and can be used to track division progression (9). However, the use of CFSE in this assay, commonly used at concentrations between 3.0 and 10.0 µM/ml for RM samples (3,10–19), must first be optimized independently, and then requires additional steps for cell labeling, and may cause significant biases that are associated with dye quality, labeling concentration and efficiency (20,21).
We describe an indirect method focused on the detection of cellular markers associated with proliferation as a substitute for the CFSE-dilution assay that we demonstrate is not affected by the biases associated with the chemical dye. Ki-67 is an intracellular marker for proliferation and a nuclear and nucleolar protein antigen present in all proliferating cells during the active part of the cell cycle: G1, S, G2, and mitosis (22–24). This antigen has been used to assess clinicopathological correlations of various tumors (25,26) and in animal models of disease (27), and is evaluated by a number of staining techniques including immunohistochemistry (28) and flow cytometry (24). A proliferation index (PI) is commonly ascribed and can be calculated by dividing the number of cells that stain positively for Ki-67 with the total number of cells in the sample group (29). We find that immunostaining for Ki-67 expression and analyzing proliferation as a function of cellular side scatter (SSC) as a percentage of total cells by flow cytometry yields similar estimates of proliferation as does the CFSE dilution assay. We validate this method in a group of sixteen RMs, four simian immunodeficiency virus-infected and twelve vaccinated, uninfected animals, and demonstrate little variability in a longitudinal and comparative studies. Furthermore, we show that CFSE labeling at commonly used dye concentrations can directly cause substantial levels of cell death during the course of the ex vivo stimulation assay. However, this phenomenon is dose-dependent and can be alleviated by further dilution while still yielding results that are correlative with staining for Ki-67. This new indirect method of assessing T cell proliferation ex vivo provides investigators with an alternative to using radioactive material or the biases associated with the chemical dye, CFSE.
We used a standard CFSE dilution assay to determine the percentage of proliferating T cells in virus-infected rhesus macaques (RM) (3). After five days in the presence of medium alone, cognate peptides or ConA, CFSE-labeled cells were immunostained, acquired on a flow cytometer, and then analyzed. CD4+ and CD8+ T cells were selected by gating on lymphocytes, live CD3+ cells, then as a positive and negative function of CD4 and CD8 expression, respectively (Supplementary Fig. 1a). Little background proliferation was observed in the control wells demonstrating that proliferation in the peptides-containing samples was specific for peptide antigen. As expected, stimulation in the presence of ConA produced robust T cell proliferation. CFSEdim cells were gated and the percentage of total CD4+ and CD8+ proliferating/proliferated cells was determined by subtracting background proliferation in the samples containing only medium from the value in the peptides-containing sample (Supplementary Fig. 1b). These data show that the widely used CFSE dilution assay can be employed to estimate the total amount of RM T cell proliferation in response to ex vivo peptide stimulation as practiced by many laboratories.
Since Ki-67 expression has previously been used to estimate PI (29), we set out to examine its relative expression in responding T cells during the proliferation assay. CFSE-labeled PBMCs were stimulated for 5 days, at the end of which they were stained for intracellular Ki-67 antigen, and then analyzed as a function of CFSE (Fig. 1a). Variable expression of Ki-67 was observed in CFSEdim T cells, as seen previously (30), and if used solely to estimate total proliferation, yielded a value that was approximately 1.7-fold lower than determined by CFSE dilution assay (Fig 1b). The lower level of proliferation detected by Ki-67, as compared to CFSE, was not statistically significant (p = 0.437), but the discrepancy between the two methods is clearly evident in Figure 1b. Also, Ki-67 was preferentially expressed in cells that had divided after five days in culture, the fraction of Ki-67+ cells being progressively higher in the populations that had undergone more divisions (30). Thus, these data demonstrate that not all CFSEdim T cells are Ki-67 positive (31) at the end of the stimulation period and that the use of Ki-67 expression alone for measuring proliferation would underestimate the true proliferative value. We next undertook additional analyses to examine the feasibility of using Ki-67 staining for the purpose of estimating proliferation.
We next examined CFSEdim T cells for expression of common phenotypic markers and their ability to diffract light by forward (FSC-A; a measure of cell size) and side (SSC-A; a measure of cell complexity or granularity) scatter. CFSEdim (dividing/divided) and CFSEbright (non-divided) T cells were graphed to look at CFSE dilution as a function of CD95 expression and data for CD8+ T cells is shown (Fig. 2a). Peptide-stimulated cells were then analyzed for the identification of highly regulated phenotypic markers commonly used for differentiating T cell subsets (Fig. 2b). Of the markers and parameters examined, CFSEdim cells were high for Ki-67 expression and the ability to diffract light to the side, which is a known marker for blastogenesis (32–35); proliferating/proliferated cells were approximately 6.5-fold higher in Ki-67 expression and 4.0-fold higher in SSC-A, as determined by gating the regulated populations (Fig. 2c). Markers not highly regulated during proliferation included FSC-A and CD95, which were 2.0-and 0.6-fold higher, respectively, as well as CD45RA and CD28, which were both approximately 1.5-fold lower in expression. These data demonstrate that for the markers examined, Ki-67 expression and side scatter capacity were among the highest regulated characters of CFSEdim, or proliferating and/or proliferated T cells.
Since expression of Ki-67 alone was a poor marker for determining proliferation, we next examined whether its expression as a function of cellular side scatter yielded a better assessment. CFSE-labeled PBMCs were stimulated and analyzed as described above and CFSEdim cells were then gated and overlayed onto CD8+ T cells as a function of Ki-67 expression and SSC (Fig. 3a). Analysis revealed that a majority of the cells were SSChigh and Ki-67 positive. Furthermore, calculation of proliferation from cells based upon their expression of Ki-67 and/or SSC showed that total SSChigh cells or total Ki-67+ and SSChigh (Ki-67/SSC) cells yielded data that was highly correlative to that as determined by gating CFSEdim cells (p < 0.01 for both SSChigh and Ki-67/SSC)(Fig. 3b). As seen in Fig. 1a, the total percentage of Ki-67+ cells alone was insufficient in estimating proliferation when compared with CFSE dilution. Also, few cells were observed in the Ki-67+/SSClow cell populations as demonstrated by a total proliferation value of less than 1%. Thus, these data demonstrate that including all Ki-67+ and SSChigh cells within a gate (Fig. 3c) provides estimates of proliferation that are comparable to those from CFSE dilution assay.
We next performed comparative studies for the assessment of proliferative responses using CFSE dilution and the Ki-67/SSC gating strategy described in Figure 3c in four SIV-infected and four vaccinated, uninfected RM animals. PBMCs initially labeled with 1.25 µM/ml of CFSE, which is well below the commonly used average of 5.0 µM/ml, were stimulated, stained, and analyzed as described above and results are displayed for T cells responding to several peptide antigens at days 0, 2.5, and 5 (Fig. 4). Similar results were observed using either method for assessing proliferation for CD8+ T cells and CD4+ T cells (data not shown) responding to two different viral antigens, SIVgag and SIVpol. Differences in proliferation values between the two strategies were minimal for samples stimulated with either peptide pools or ConA. While the expression of Ki-67 was observed to be elevated in infected animals (graphs at left) and is consistent with a chronic immune activation profile (36,37), values decreased by day 2.5 and were dramatically reduced in the samples that were incubated in medium alone on days 2.5 and 5 (data not shown). Therefore, these data show that gating for expression of Ki-67 and side scatter in peptide- or ConA-stimulated samples yield a proliferation value that significantly correlates to that obtained by CFSE dilution assay as measured 5 days after stimulation in either SIV-infected (peptide: r = 0.978, p=0.022; ConA, r = 0.992, p=0.008) or vaccinated, uninfected (peptide: r =0.970, p=0.030; ConA, r =0.990, p=0.010 ) animals. Altogether, these data validate the use of Ki-67 staining as a semi-quantitative surrogate for CFSE dilution in estimating T cell proliferation in rhesus macaques.
We next examined the contribution of CFSE-dye to in vitro culture conditions and the estimation of proliferation ex vivo. We hypothesized that dye-mediated cytotoxicity may be a contributing factor to this variation since (1) CFSE-mediated cytotoxic effects have been described (20,21), (2) Ki-67 negative and SSClow populations were diminished in ConA stimulated wells (Fig. 3a), and (3) experience from our laboratory suggests that using CFSE-labeling in vitro for stimulation periods longer than three days results in noticeable cell death (unpublished observation). RM PBMCs from four infected and four vaccinated, uninfected animals were either labeled with CFSE (1.25 µM, which is well below the commonly used average concentration of 5.0 µM/ml (3,10–19)) or not treated, and then incubated in medium alone for 5 days (Fig. 5). Samples were harvested from culture on days 0, 2.5 and 5, stained, and acquired by flow cytometry immediately after the designated incubation period. Results are shown as percentage of dead cells (as measured by viability dye incorporation) for each animal over time (Fig. 5a). Substantial cell loss was observed in the CFSE-labeled samples when compared to unlabeled cells (Fig. 5b). CFSE-labeling accounted for a significant amount of cell death on Day 5 (p<0.01), approximately 19% more cell death, than did samples with no dye. There was no difference in the average cell death between the samples from infected and uninfected animals (data not shown). Altogether, these data confirm previous reports (20,21) and show that labeling with CFSE dye during long-term in vitro stimulation assays can account for marked cytotoxicity.
To further characterize the effects of CFSE labeling during the ex vivo proliferation assay, we next examined cytotoxicity and Ki-67 data from cells stained with lower amounts of dye. Cells from eight additional uninfected and vaccinated animals were labeled with 0.125 µM/ml of CFSE, a one log lower dose than reported in Figure 5, and stimulated for 5 days in the presence of media alone, cognate peptides, or ConA (Fig. 6). Labeling of cells at a lower concentration (0.125 µM) resulted in less incorporation of dye at the end of the stimulation period in media alone compared to samples labeled with 1.25 µM and those with no CFSE, as determined by flow cytometry (Fig. 6a). The percentage of cell death resulting from each condition is displayed (Fig. 6b) and shows that cytotoxicity is dose-dependent; cell death occurring from labeling with 0.125 µM is minimal (32.1%) when compared with that from unlabeled cells (32.5%), which are both significantly lower than that from labeling with 1.25 µM of CFSE (52.5%) (p<0.01). While data from CFSE gating and Ki-67 gating were highly correlative at a CFSE dilution of 1.25µM, we next examined whether this correlation was maintained at the lower concentration. Proliferating/proliferated cells labeled with 0.125 µM CFSE were gated based upon CFSE or Ki-67 staining and proliferation levels were found to be highly correlated when either stimulated with cognate peptides (p<0.01) or with ConA (p<0.05; Fig. 6c). Altogether, these data show that CFSE-mediated cytotoxicity is dose-dependent and can be diluted to minimize cell death while yielding data that is still correlative between CFSE gating and Ki-67 gating.
The CFSE dilution assay is popularly used for assessing cellular proliferation, monitoring the number of cell divisions during proliferation, and the examination of the relationship between proliferation and differentiation. The proliferative response can generally be examined in different ways depending on the application. Quantitative methods that consider the differential CFSE fluorescence halving for grouping the dividing cells into different daughter cell generations and then used to evaluate the cell population in each generation separately (38) are useful when also analyzing the expression of dynamically-regulated markers as a function of division (3,39). However, when simply assessing T cell proliferative capacity as a measure of vaccine-induced immunity, it may be easier to apply a semi-quantitative method since it does not consider different degrees of CFSE fluorescence halving in proliferating cells, but alternatively, all the cells with fluorescence lower than that of the undivided parent generation are considered as one responding population (3,12,15,17–19). Thus, proliferative capacity as measured by total T cell proliferation is an accepted method especially in comparative studies where, for example, ex vivo responses are elicited against cognate peptides to estimate the quality of immune potential in immune diagnosis, pathogenesis, or vaccine studies.
Using Ki-67 in assessing proliferation as a surrogate for CFSE, as described here, does not allow for the quantitative evaluation of cell division number, or for the examination of cell differentiation as a function of division (15,30). This assay is proposed as an indirect and semi-quantitative method to be considered as an alternative to using CFSE for estimating proliferative capacity. We find that it is a viable method based on the properties of increased and detectable expression of both Ki-67 antigen and SSC, which are both known to be associated with blastogenesis (34). As viewed in Figure 3, the bivariate distribution of Ki-67 antigen and side light scatter may distinguish the populations of cells responding to antigenic stimulation; all Ki-67-positive cells represent actively dividing cells and all SSChigh cells represent cells of higher granularity, or intracellular complexity (33). Cells that are CFSEdim and Ki-67-negative likely represent cells that have subsequently stopped dividing and entered into quiescence. However, these cells are also SSChigh, which may be a consequence of a number of factors associated with cell division or the assay itself. For example, these cells may have recently divided, but have not yet fully reorganized the cytoplasm and DNA and are thus still SSChigh. Or perhaps they retain remnants of an altered metabolism that was used during division which may persist for longer periods or at least the duration of the in vitro assay following the introduction of quiescence (35). Indeed, complexity as a measure of SSC consists of protein, RNA, and DNA content, and the texture and refractive index of intracellular particles (32). Reflection from small particles is the main component of side scatter and because SSC-A is the time integration of side scatter, it is proportional to the number of particles in the cell, and as a consequence, is proportional to cell volume (40). This phenomenon was not a consequence of chronic SIV infection since it was also observed in antigen-experienced, uninfected animals. Thus, distinct populations could be segregated based upon the expression of these two markers; Ki-67 positive, but SSClow cells likely constitute a minute percentage of CFSEdim cells that are just re-entering interphase (G1 or early S phase) following subsequent quiescence; double positive cells are most likely cells that are actively dividing in late S phase, G2 and mitosis, since the content of Ki-67 nearly quintuples during late S phase (41). We validated this method using PBMCs from four SIV-infected and twelve vaccinated, uninfected RMs since ex vivo antigen-driven proliferation is readily achieved in these models (42). Since gating in this manner yields proliferation values that are highly comparable with values from the CFSE-based assay, this method can be used as a surrogate of CFSE for assessing cellular proliferation, but not cellular differentiation as a function of division.
Detrimental effects of CFSE on lymphocyte function and viability is not generally considered even if it has already been described (20,21,43). Lašt’ovička and colleagues (21) show that CFSE-labeling of human PBMCs accounts for up to 50% cell death using dye concentrations ranging between 5 to 10 µM, which are commonly used for macaque cells (10 µM (11,15,16,18), 5 µM (3,10,12,13,17–19), and 3 µM (14)) and is the recommended range by the manufacturer for long-term staining (more than three days) or for the use of rapidly dividing cells; we used a concentration of 0.125 to 2.5 µM in this report but recommend that each lab optimize CFSE concentrations independently for each application. Indeed, we also observed the cytotoxic effects of this intracellular dye at a concentration (1.25 µM/ml) well below the average value of 5.0 µM/ml for macaque cell proliferation studies which reached up to 19% cell death by day 5 of the ex vivo culture period. While it is unknown as to the effect extensive cytoxicity may have on the assessment of proliferative responses, which may or may not target specific dividing or quiescent cell populations, it is clear that increased amounts of cell death due to CFSE biases may negatively affect the validity or physiological relevance of the assay. For instance, higher amounts of free cellular debris in the sample wells may non-specifically sequester peptide antigen, or may even interfere with cell function, metabolism, or proliferation directly. However, it is most likely that a number of factors in combination with CFSE-labeling in this assay may facilitate cell death such as assay duration, quality or concentration of dye reagent, and the specific species of cells being examined in long-term tissue culture. Since there is an indirect relationship between cellular CFSE content and proliferation, non-divided cells retain the highest levels of the chemical dye and therefore may be more susceptible to dye-mediated cytotoxicity, while actively dividing cells decrease the total amount of CFSE by half upon each division. Furthermore, it is often implicitly assumed that labeled cells are representative of the entire peripheral lymphocyte population and therefore that the kinetics obtained from labeling experiments pertain to the whole lymphocyte population (7). However, it should be presumed that in cases where CFSE-mediated cytoxicity is abundant, proliferation values may be artificially increased due to the potential biased depletion of non-dividing cells which would thus help to increase the value of the ratio of dividing cells over non-dividing cells. Thus, little benefit can be anticipated from CFSE-associated phenomenon in the determination of ex vivo T cell proliferative responses and it is recommended here that the dye be used at concentrations which minimize these effects.
One way to potentially mitigate this CFSE-associated toxicity is to further dilute the amount of CFSE used, as recommended by numerous protocols. Indeed, we found that using the CFSE dye at a lower concentration of 0.125 µM eliminated dye-mediated cytotoxicity and that CFSE proliferation data were still correlated with data obtained by Ki-67 gating. However, it should be emphasized that further dilution may limit the range of detectable cell division (44), which is particularly critical when assessing the dynamic regulation of cellular markers as a function of cell division. Irregardless, the Lašt’ovička et al. report also shows that the use of CFSE in general (at concentrations ranging from 37 nM to 10 µM) impairs the proliferative capacity due to the decreased viability of proliferating cells in a concentration-dependent manner. They go on to demonstrate that this may be a result of substantial modulation of cellular activation molecules (increased human activation markers such as HLA-DR and CD27, while decreasing CD25 and CD71 during mitogenic stimulation) and that the use of CFSE in diagnosis of severe cellular deficiency yields a high proportion of false positive results. Thus, assessing proliferation in the absence of CFSE as reported here should help to eliminate these putative biases including the artificial inflation of proliferation values, depletion of nondividing cells or others, and those that may help to even decrease proliferation values, impair proliferative activity, and alter expression of activation markers (21). While it remains to be determined how the absence of CFSE will affect estimation of T cell proliferation ex vivo, it is certain that the assay will be more physiologically relevant and devoid of the biases associated with this chemical dye.
CFSE- and Ki-67/SSC-determined proliferation was comparable after 2.5 days in culture in both SIV-infected and vaccinated, uninfected animals. However, as seen in Figure 4, greater amounts of T cell proliferation were observed at day 5 than at 2.5, which could be up to seven-times higher for some animals. Furthermore, higher Ki-67 values were observed in SIV-infected animals than in vaccinated, uninfected animals and is consistent with previously observed data (36,37). These elevated values in chronically infected animals subsided by day 2.5 and remained low thereafter in non-proliferating control samples that were stimulated with medium alone. Therefore, the method presented here is informational for assessing antigen-specific proliferative responses in antigen-experienced uninfected animals, and in chronically infected animals after at least 2.5 days in culture.
While the scope of this study was focused upon a study length of five days, it may be possible that informative proliferation values from Ki-67/SSC gating may be achieved at either earlier or later time points, based upon the desired application. For example, studies examining proliferation kinetics should consider that shorter stimulation periods will decrease assay length and therefore lower the amount of time required for making an immunodiagnosis in the clinic. On the contrary, longer periods of stimulation, which may be possible due to less dye-associated cytotoxicity, may yield additional information regarding the capacity of T cells to actively proliferate for prolonged periods. In this way, proliferation data may be supplemented with an index value of actively dividing (Ki-67+ cells) over recently, but not actively dividing (Ki-67-negative/SSChigh), which may reveal information about the quality of the proliferative response where higher indexes indicate superior capacities for prolonged proliferation. It would be most useful to also perform these analyses using other stimuli such as whole protein antigen and in comparison to other accurate and sensitive cell-cell culture techniques such as 3H or BrdU incorporation. Additionally, it is likely that this scheme can be applied retroactively for determining proliferation in studies that have incorporated Ki-67, and not CFSE, into the analysis of samples that were cultured with antigen.
In conclusion, we show an alternative method for assessing proliferation ex vivo. Unlike standard CFSE-dilution assays in which the cells are chemically modified for the duration of the assay, staining for Ki-67/SSC it is not associated with the biases of cell dyes, such as cytotoxicity, which we confirm here is dose-dependent (20,21,44). Thus, using Ki-67 expression for estimation of proliferation upon completion of the assay obviates the cellular manipulation using CFSE dye at any concentration, which physically interacts with various components of cells including cell membranes, surface glycoproteins, enzymatic systems, and DNA. Our validation of the method showed that its measurements correlated well with values acquired using CFSE, even at lower dye concentrations when cell death does not occur, but that it may represent a more physiologically relevant means of estimating immune function during immunodiagnostic, pathology and vaccine studies, since it is excluded from potential biases associated with chemical modification.
Rhesus macaques (Macaca mulatta) of Chinese or Indian origin were used in this study and were housed at the Bioqual facility in Rockville, MD or the Tulane National Primate Research Center in Covington, LA, respectively. These facilities are accredited by the American Association for the Accreditation of Laboratory Animal Care International and meet National Institutes of Health standards as set forth in the Guidelines for Care and Use of Laboratory Animals. Chronic SIV infection in Chinese animals (SIV1 - 4) was established using i.v. inoculation with 100 MID50 (50% macaque infectious dose) of highly pathogenic SIVmac251 and all were infected and reached peak viral loads by wk 2 and set point by wk 10. Studies were performed during late-stage chronic infection (about 80 weeks post-infection) since T-cell turnover is increased (42) and all animals had detectable levels of virus. Indian animals (VAX1 - 4) and Chinese animals (VAX5 – 8) were uninfected and immunized with a DNA vaccine at weeks 0, 6, 12, and 18 with 3.0 mg of pSIV Gag DNA, 1.5 mg of pSIVpol DNA, and 1.5 mg of pSIVenvDNA. The latter animals (VAX5 – 8) were then boosted with adenoviruses expressing SIV Gag and Pol at months 15 and 16. The remaining animals (VAX9 – 12) were primed with adenovirus at months 0, 1, and 6, and then boosted with DNA at months 12, 13, and 22. DNA was formulated in sterile WFI with 1% wt/wt poly-L-glutamate sodium salt and delivered in 2 separate sites into the quadriceps muscle in a total volume of 0.75 mL per injection (1.5 mL total per animal) followed by in vivo electroporation using the constant current CELLECTRA® device (Inovio Biomedical Corporation).
Isolation of RM PBMCs has been described proeviously (3). Blood from RMs was collected in EDTA tubes, shipped same day, immediately overlayed on a Ficoll-Paque PLUS gradient (GE Healthcare, Piscataway, NJ), then centrifuged for 25 min at 2,400 rpm. PBMCs were collected from the interface and residual red cells lysed using ammonium chloride-potassium (ACK) lysing buffer (Lonza, Walkersville, MD). Cells were washed twice with Hank’s Balanced Salt Solution (HBSS; Invitrogen, Grand Island, NY), counted using a 0.4% (w/v) Trypan Blue Solution (Mediatech Inc., Herndon, VA), resuspended in complete R10 medium (RPMI 1640 (Invitrogen) containing 2 mM L-glutamine, 100 IU/mL penicillin and streptomycin, 55 µm/L β-mercaptoethanol and 10% fetal bovine serum (FBS)), and incubated at room temperature (RT) until further use.
The T cell proliferation assay using CFSE labeling of RM cells has been described (3). Briefly, freshly isolated RM PBMCs were stained with the Vybrant® carboxyfluorescein diacetate, succinimidyl ester (CFDA-SE) (which is colorless and cell permeable, unlike CFSE, and is cleaved once inside the cell to yield fluorescent CFSE) Cell Tracer Kit (Invitrogen) per the manufacturer’s instructions. A range of CFSE dilutions (0.125 to 2.5 µM/mL stock of CFDA in PBS) was used to label 10×106 cells in this report, which is lower on average than published data for macaques which ranges from 3.0 to 10.0 µM (3,10–19). Labeled cells were washed three times in PBS then resuspended in R10 medium. One million cells per sample per well were added to a 96-well U-bottomed plates in a total volume of 200 µl of R10 medium. Cells were stimulated using either pooled peptides from SIVgag, SIVpol, or SIVenv (consisting of 15-mers overlapping by 11 amino acids; NIH AIDS Research & Reference Reagent Program, Germantown, MD) or 5 µg/mL ConA (Sigma, St. Louis, MO) for 5 days in 5% CO2 at 37°C. After three days, cell cultures were fed by removing 50 µl of the culture medium and replacing it with 100 µl of fresh R10 medium. At the end of the stimulation period, cells were washed in PBS for subsequent immunostaining and polychromatic flow cytometric analysis. Total proliferation is the percentage of CFSEdim or Ki-67/SSC-gated cells per T cell subset and then background subtracted for that observed in medium alone samples.
Immunostaining of PBS-washed RM PBMCs was performed using LIVE/DEAD® Fixable Violet Dead Cell Stain Kit for 20 min at RT (Invitrogen) per the manufacturer’s instructions and then washed once with Dulbecco’s PBS (Invitrogen). Cells were then stained for 30 min at 4°C with PBS containing 1% FBS (1% PBS) including the following Abs: PerCP-Cy5.5-conjugated anti-CD4 (clone L200; BD Biosciences), AmCyan-CD8 (clone SK1; BD), PE-Cy™5-CD95 (clone DX2; BD), ECD-CD28 (clone CD28.2; Beckman Coulter, Fullerton, CA) and Qdot® 605-CD45RA (clone MEM-56; Invitrogen). After three washes with PBS, intracellular staining was performed using the BD Cytofix/Cytoperm™ kit and Abs APC-Cy™7-CD3 (clone SP34-2; BD) and either APC- or PE-labeled Ki-67 (clone B56; BD) in perm/wash buffer for 30 min at 4°C. Following staining, cells were washed twice in perm/wash buffer and fixed with 1% paraformaldehyde. Data was collected using a LSRII flow cytometer (BD) and a sample of CFSE-labeled cells that was incubated with medium alone was used as a control for compensation. Flow cytometry data was analyzed using FlowJo software (Tree Star, Ashland, OR) and cells were gated on singlets using FSC-H by FSC-A and antigen-experienced CD95+ cells.
All values are reported as the mean ± SEM. Correlation coefficients between proliferation determined by Ki-67/SSC and CFSE were calculated according to Pearson. Analysis of cell death was completed using a 2-tailed Student’s T-test (Figure 5) and by ANOVA (Figure 6B). Fisher’s Least Significant Difference test was used to adjust for multiple comparisons (Figure 6B). Statistical significance was assumed at p≤ 0.05. All statistical analysis was carried out using the Statistical Package for the Social Sciences (SPSS, Chicago, IL).
We would like to acknowledge M.A. Kutzler, A.J. Sylvester, N. Kathuria, J.D. Boyer, A. Dai, G. Silvestri, N.R. Klatt, A.M. Ortiz and members of the Weiner laboratory for significant contributions and/or critical reading of this manuscript.
COMPETING FINANCIAL INTERESTS
The laboratory notes possible commercial conflicts which may include advising, consulting, or collaborations with Wyeth, Inovio, BMS, Virxsys, Ichor, Merck, Althea, J&J, and Aldeveron. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. No writing assistance was utilized in the production of this manuscript.
1Support from the National Institutes of Health AIDS Research and Reference Reagents program (NIH-ARRR and University of Pennsylvania Centers for AIDS Research (CFAR), is also acknowledged. This work was supported in part by grants from the NIH to DJS (T32 AI070099) and DBW.